Sir, - In a typically vague column (November 17th), William Reville makes several claims about the heritability of such apparently genetic traits as religious beliefs, political inclinations, beauty and, of course, IQ. If Dr Reville really thinks that the tendency to vote Fianna Fail is a non-additive genetic trait, I doubt if I can convince him otherwise. I have to admit that it's a tempting hypothesis. However, his assertion that "intelligence has proved to be 60 to 70 per cent heritable" indicates that Dr Reville's tendency to play fast and loose with concepts and evidence has gotten the better of him once again. The "growing consensus" described is in reality difficult to find. Two points are in order. First, he jumps directly from measures of IQ to claims about intelligence. Debate in this area has gone on long enough for us to know that such a jump is almost certainly unwarranted. The fact that IQ is the single most heritable trait in twin studies - presumably ahead of many more directly measurable physical attributes - might make us more rather than less suspicious about the reliability of the IQ measure. James Flynn's work, showing that test scores have been rising steadily over the years for no good reason, is important here.
Second, and more important, evidence from twin studies is and always has been problematic. Difficulties range from outright fraud (in the case of Sir Cyril Burt's work) to more mundane, but still serious, methodological problems. In particular, many twin studies rely on the dubious assumption that when twins are separated they are distributed randomly across environments. I refer Dr Reville to Prof Howard Taylor's The I.Q. Game for evidence of just how questionable this assumption has been in many studies. Supposedly "separated" twins have been found living with a different branch of the same family, for example, or in the same neighbourhood as their sibling. In short, Dr Reville is right to say that genetic and environmental characteristics interact in complex ways. But he is wrong to imply that the specific claims he makes about the heritability of intelligence represent a scientifically determined consensus. - Yours, etc.,
Kieran Healy
Department of Sociology, Princeton University, New Jersey.